@article{ref1,
title="Response to "Comment on: Machine learning for understanding and predicting injuries in football"",
journal="Sports medicine open",
year="2024",
author="Majumdar, Aritra and Bakirov, Rashid and Rees, Tim",
volume="10",
number="1",
pages="e85-e85",
abstract="We acknowledge the Letter to the Editor by Bullock and colleagues [1] regarding our article &quot;Machine learning for understanding and predicting injuries in football&quot; [2], and appreciate the opportunity to respond. In our Leading Article [2], we outlined the topics of sport injury and machine learning, before describing examples from the literature that had used machine learning to examine the workload-injury relationship in football. Our aim was to &quot;aid readers both from sport science and machine learning communities in their understanding of sports injury articles employing machine learning&quot; (p. 2) [2]. We concluded: &quot;the myriad ways machine learning can be employed can also lead to difficulty in synthesising the current research evidence into an overall, unified, conclusion. Indeed, there remain questions as to the utility of these models for real-world application&quot; (p. 8) [2].   Given the above, we were confused as to the content and purpose of the letter [1]. The letter either (a) raised points with which we have not disagreed, backed by citations to the letter authors' own work, (b) made general observations about machine learning, or (c) countered points we never made--strawman logical fallacies.   The letter made five key points [1]. The first point--that we claimed the models that we reviewed in our Leading Article [2] were &quot;quite sound&quot;--is untrue. The letter authors [1] noted...   Keyword: Soccer<p /> <p>Language: en</p>",
language="en",
issn="2199-1170",
doi="10.1186/s40798-024-00751-3",
url="http://dx.doi.org/10.1186/s40798-024-00751-3"
}